PDA

View Full Version : Splints router build log



Splint
12-24-2005, 05:18 AM
Hi Gang,
for quiet some time now I have been working on and off on my first router. Over the xmas break I have been working on it full time and hope to have it completed and working in the very near future.
My primary reason for building a cnc router was to be able to quickly and accurately make my own foundry patterns, and with that in mind I designed it with a reasonably large z travel (280mm) and will be using an electric die grinder with an extended "snout" to have a better ability to work in confined pockets. A secondary reason for the long z travel was to build it with the ability to use it as a starting point for a four axis and then a five axis machine at a later point in time.
It will be a servo driven system using Electrocraft E543 motors with 1024 cpr encoders and will use Rutex 2010 drives and Isel rolled ballscrews. I hope to be able to get rapids speeds of about 100ipm.
The table is 1800 long (1550mm x travel) and a bit over a meter wide (990mm y travel).
Anyway, nuff talk for now, here are some pictures of the construction of the table.

ynneb
12-24-2005, 05:26 AM
I just got one word to say to you.
ABOUT TIME ! ( actually two words )

Splint
12-24-2005, 07:32 AM
Thanks for the encouragement yennb....I hope you realise that in order to make space to store this machine once completed I actually had to buy a house, then once I had the house I took full advantage of the silly amounts of overtime going at work and was regularly doing 90 hour weeks so no time for building a router...
Anyway getting back to the story, the x and y axies are based around a 200 by 250 by 20mm alloy plate which has 10mm angle plate bolted to it. The angle plate will be the mounting point for v blocks which will run skate bearings, the whole assembly will roll on 3/4" chrome plated hydrolic ram bar, and there will be plates bolted to the end of the angle plates to control any spreading effectof the angle plates while in opperation.
As you can see in the picture, prior to drilling the holes in the plate I laid the angle plates in their positions and marked the holes with a texta so that if I misread the dro or the plans I would pick up the error prior to drilling holes.

Splint
12-24-2005, 07:55 AM
Here are some more pictures of the construction. One of the side plates is shown, there will be a plate run across the bottom, below the table where the x axis ballnut will mount (centerally).
To the left of the side plate is the two rails which will sit in vees machined in an aluminium plate. This will be used for the z axis, which will be quiet narrow, marginally narrower than the die grinders case to allow for movement into confined spaces.
The y axis plate is mounted in the vise on the mill and has had the bolt holes drilled and tapped for the rails to be mounted in.
The other componants are the mounting brakets for the x axis v blocks, the lower section of the z axis where the snout of the die grinder locates and the side braces to assist with lateral rigidity of the side plates.

Splint
12-24-2005, 08:30 AM
The electrical system is the remains of a small mig welder which I got from a guy who repares welders, he took of the wire feed, leads and gun and i got the rest with a few big caps. I put a few more turns on the secondry winding to get the desired voltage and I was in business. Allthough as far as power supplies go, this is overkill, even if I go for a full 5 axis machine and all motors are pulling maximum amps at the same time there will still be plenty in reserve. The main reason I went this way was because of the price (far less than the cost of new componants if I were to build one of my own) and the simplicity of having a working system ready to go.
The biggest issue for many Aussies is there is such limited availability of second hand cnc gear and the price of such gear is high (that's if you can even find what you want) fortunately I picked up my servos at a reasonable price when I was in Singapore at Kaichin Electronics who deal in used industrial electronics and hardware.
As mentioned earlier the servo drives are Rutex 2010s and I opted for an optoiscolated motherboard to protect the pc. Unfortunately I accidentally hooked up the power supply wires to a board reverse polarity and let the smoke out. The pc I got from a computer swap meet, it's a pentium 2 gig.

Ken_Shea
12-24-2005, 08:52 AM
It is nice to get some spare time I bet, 90 hours does not leave much left, however it looks like you are making up for it now, looks great.

I hope you find time to share some close up pictures of the Rutex 2010's and associated connections. That is a ways away for me yet and I have been reading up on it to get some mental understanding but actual pictures would be a real help and time saver.

Ken

Rodm1954
12-24-2005, 11:08 AM
Splint
Keep the posts coming - fantastic work.

I know what you mean about limited materials and high prices. I have been sourcing bits from Singapore as well - it seems like the best place for used gear over this side of the world.

Splint
12-24-2005, 08:36 PM
The pulleys are AT5 type which is a more precision version of T5 pulleys (less backlash between the belt and pulley) the belts are synchroflex brand, not sure where they are made, possibly here in Australia. The energychain is Kabelschlepp microtrac http://www.kabelschlepp.com/miva/merchant.mv?Screen=CTGY&Store_Code=K&Category_Code=CHCS Igus also has a good range of energy chain but Kabelschlepp was cheaper.

Splint
12-24-2005, 08:45 PM
Here are some close ups of the Rutex system. The first image is the optoiscolated motherboard with a 2010 drive plugged in. I bench tested it like this and used the Rutex tuning software, if you follow the instructions it's pretty easy.
The next image is a closer look at the 2010 drive.
The last image is the single ended line driver. The encoder is wired to the screw terminals (pos, neg A and B) and then a network cable runs between the line driver and the 2010 drive.

Splint
12-25-2005, 05:20 AM
This image shows the ballscrew which is 16mm with a 2.5 mm pitch which will be the spec for the y and z axies. The x will use a 16mm diameter with a 5mm pitch to halve the rpm of the screw to reduce the whipping effect which will be caused by its considerable length.
The ball nut, seals and housing are also shown. The housing has two grub screws in it, one at the end to lock the nut into the housing and another at the centre to tension the nut. The nut has a cut out section which has a shim device in it, once the screw is in the nut the tensioner screw is tightened until the shim bottoms out and once the nut is lubricated through the grease nipple on the nut mount it is ready for use. The shaft through the nut is a dummy shaft which remains in place until the nut is fitted to the shaft, this prevents the balls from dropping out.

Prboz
12-25-2005, 07:05 PM
Hey Splint, When you said you where building your machine in your PM to me I didn't expect what I found in your log. Don't beleive in starting small hey :).

Will keep uptodate of your progress, will have to go back to my humble shed and keep building my TOY.

Later

Prboz

Prboz
12-25-2005, 07:07 PM
Do you have any singapore web sites for equipment?

Thanks

Rodm1954
12-25-2005, 07:56 PM
A search found this

http://cnczone.com/forums/showthread.php?t=3125&page=5

I have sent you a PM with more details.
Hope it helps

Splint
12-26-2005, 06:41 AM
Hi Prboz,
I guess "big" is relative, you should see some of the mosters used in the marine and aviation industries, I'm talking work envelopes of 50 by 12 by 8 meters, that's big.
Ebay has several stores in Singapore who specialise in cnc hardware but that's all I know of. Kaichin, to the best of my knowledge dose not have a web site.
Here is an image of the die grinder I will be using. The base of the z axis is also shown, I will have to modify it by changing into a clamp where the snout of the die grinder locates.
I also drilled and tapped the z rails. The alloy part (not shown) that the rails mount to is not thick enough to support bolts coming in from either side so I have made it so that the bolts come from one rail, pass through the alloy and screw into the rail on the other side. I used an end mill to cut a recess for the head of the unbreako bolts.

Splint
12-27-2005, 06:40 AM
I drilled the y axis rails today and did a trial fitment to the plate. I was machining recesses for the heads of the bolts such that they sit inside of the circumfrence of the rail and the end mill went blunt, a combination of too high rpm and the hard chrome surface. I didn't have a spare handy and all the industrial suppliers are still closed for xmas so I just used a 9mm drill and didn't drill too deeply, I'll rebore them when I can get hold of an end mill.
I measured the width variation of the upper and lower rails on the y plate with a digital vernier and was pretty happy with the results. The total variation was 0.05mm, the lower limits were of course at the bolt points and the upper limits were mid way between the bolts. I should be able to get the lower limits more uniform by adjusting the bolt torques, that was the reason i designed it so the rail will not bottom out in the channel. When the skate bearings (being set to match the lower limits) pass over the sections mid point between the bolts the rails should squeaze down and closely match the lower limits so the overall variation should be minimal.
I need to buy some bolts long enough to mount the z axis rails before i can check the variation on them, I'll be doing that shortly.

Splint
12-27-2005, 07:39 AM
Some more detail of the Rutex system.
The wiring on the power terminal: the blue wire (which should be fused) runs to the power supply positive terminal, brown to the power supply negitive terminal, black and white to the motor terminals, polarity dose not matter on the motor terminals as that can be changed with the tuning software.
The red cable is rj45 network cable which runs to the single ended line driver (which is wired to the encoder). The 15 pin connector is used to power 2 circuits on the motherboard and to close 2 circuits (e stop and out1/spi) to allow the tuning to take place. The printer cable runs to the pc.
There are several different Rutex tuning software packages which essentially perform the same function, the grapical user interface is the main difference, this one is the r2xtunevb6 package.
This is the graph which the tuning software produces, the instructions shows an ideal graph and the objective is to adjust the parameters to get your graph to look similar to the ideal graph. The motor can jump around and oscilate violently depending on the parameters, I would tune with the motors disconnected from the leadscrews to avoid possible damage, there are some notes on safety when tuning in the instructions.

Splint
01-04-2006, 08:46 AM
Hi guys,
I'm back again after a few days holiday. I machined the bearing mounts for the x axis today. I made a jig to mount the blocks to so they could be machined as a set. Doing it this way is more accurite than mounting them in a vice as they can move too easily when tightening the vice.
The first picture shows the blocks mounted to the jig with the jig bolted to the table and the head tilted on a 45 degree angle to cut the faces of the mounts.
The second image shows an edge finder in use. For those who are not familiar with edge finders they are a probe like device with a spring loaded ball on the end, you slowly move the table intil the ball touches the surface of the job and the electrical circuit is closed which makes a red light come on and a buzzing noise. When it is at the point where the slightest movement can make or break the circuit you set the DRO to either + or - 5mm (half the diameter of the ball) depending on the direction the table is moving and you have your edge point. After that remove the edge finder and put in the cutting tool and start machining.
The last image shows the finished bearing mounts for the x axis which will run on 1" diameter shafts. Behind the bearing mounts is the jig (upside down).

Splint
01-06-2006, 06:01 AM
I didn't have a very productive day today, I got bogged down with some other stuff but I did manage to get the y bearing blocks made and did a trial fitment. I tightened the angle brackets flush to the edge of the backing plate and bolted on the bearing mounts and put in the y plate, with a slight spring tension as the rails passed throught the bearings the motion was smooth and completely free of play, perfect. Here is an end view and back view of the y axis.

Splint
01-07-2006, 05:04 AM
I mounted the rails of the z axis and did a width variation check today. There was .09mm variaton across the rails showing it to be unparalel so I disassembled it and wrapped a sheet of emery paper around one of the rails and rubbed the high points and got it down to a very respectable .02mm variation, the narrow sections were at the bolt points and the wide sections were mid way between, as expected.
I machined the z axis bearing mounts. The first image shows setting up prior to machining, I mount the parralels to the side of the table and then pull the job back against the spacer bar and torque the jig bolts, that way is is true to the table.
The next two piccies are of the router. We've been getting a few visitors and trying to explain what I'm building is difficult so I pinned up a couple of pictures on the wall. The final job will have 4 1" by 2" rectangular tubes running across the frame to support the mdf table top.
I tried to mount the z axis and the rails were slightly too wide for the bearings, I'm not sure what went wrong but some elongated holes in the angle plate sorted it out. Moving the z up and down its travel I noticed a very slight bump which turned out to be a tiny dint in the surface of the rail right where the bearing runs. I dont think it will be a major issue but the rail can be easily replaced if required.

Splint
03-13-2006, 03:57 AM
Another update (just for you Rod ;-) ),
I modified the bottom plate on the z to take the router and made a bearing housing for the leadscrew which also ties into the main member to control the hinging effect which will be generated as the gantry moves on the x axis.
The four rings are adaptor plates for the servo motors so the side loads produced by the belt tension acts on the boss on the servo motors rather than the tiny mounting bolts. I will only be using three motors on the router initially but 4th and 5th axies are on my list of things to do at a later point in time.
The two rectangular bearing mounts in the side view image are for the y leadscrew. The "top hat" bearing housing is for the z leadscrew.
Rather than machine the ends of the ballscrews to take the pulleys I intend machining adaptors which will mount to the ends of the leadscrews and the pass through the bearing housings then the pulleys and then have a threaded section for a nut, that way if there are any machining errors on the adaptor it's no big deal, a replacement can be easily made rather than an expensive replacement of the ballscrew.

Cheers
Splint

BALLSCREWPRO
03-18-2006, 01:40 PM
Store in Singapore, My machine shop is selling CNC parts, Website: www.pro.com.sg
No other places aound where you can find so much cnc parts, Imagine 2000 sq ft.
stack up from the floor to the 12' ceiling. I have stepper and servo driver setup to test motors to ensure running conditions. At the same time you can see how i retrofit
my machines and enrich your ideas, Remember, Knowledges is your bank account.
If happens to pass by Singapore, email me : chan@pro.com.sg
Cheers

Splint
04-20-2006, 07:06 AM
I am part way through making these housings for the x axis drivetrain. The housing in the background is a 50% reduction ratio as the x leadscrew has a 5mm pitch and the y and z have 2.5mm pitch.
The housing in the foreground is the motor mount and will be fixed in position on the cross brace at the end of the table. The 50% reduction assembly will be mounted between the motor mount and the leadscrew, it will be mounted on elongated bolt holes so it can be used as a tensioner for the two belts.

svenakela
04-20-2006, 11:46 AM
Wow!

Splint
06-10-2006, 10:17 PM
Some more images.
The first and second images are of the pulleys being checked for run out prior to boring them.

The third image is the pulleys after they have been bored.

The fourth image is the x axis components (background) and the y and z axis components (foreground).

The fifth image is a closer look at the z drive pulley extension.

Rodm1954
06-11-2006, 07:25 AM
Splint
Good to see your machine is progressing. It must be gettig close to a 10 carton job by now. :)
Great work and keep the posts coming.

Splint
10-02-2006, 08:17 AM
Progress has been slow but I hope to have it finished soon. Here is a picture of the y axis assembly. I will be adding more alloy plate to form a box section on the y axis as there is plenty of rigidity in the virticle plane but not enough in the horizontal plane.

Splint
02-03-2007, 06:38 AM
Establishing a level surface for the table to mount on proved more difficult than expected. I used 50 by 25 rectangular tubing as support braces for the table to rest on. I assumed that they would be flat enough as is but running a straight edge across the table showed more deviation than I felt was acceptable. I mounted the rails in the mill and took a light cut off the top and rechecked them in the machine, still there was some variation in the surface which I was pretty surprised to see but attributed to the way the rails were mounted in the mill. I re mounted them in a way which would give minimal warpage or preload on the parts and remachined them. Again, there was excessive variation in the surface. I was running out of metal to machine off the rails so I resorted to using body filler (bondo, I think the Yanks call it) and a straight edge. I coated the alloy straight edge with oil so the body filler wouldn't stick to it and ran a bead of filler along the rails. Using welding magnets to keep the straight edge located I pushed the straight edge into position and troweled off the excess filler. To cut a long story short, the body filler did a much better job of sticking to the oily alloy than the steel rail so after some experimentation to see what filler wont stick to I coated the straight edge with grease which turned out to be a success. Pictured is the straight edge with the filler setting.

Splint
02-03-2007, 06:46 AM
I partially assembled the gantry for the first time so I could run a dial indicator along the edges of the table to see if the x rails are running true to the table top. This is the first step in truing the machine up, I can see this is going to be a slow tedious process.

Greolt
02-03-2007, 07:11 AM
She's really starting to look like something now Splint. :)

Splint
01-04-2008, 08:29 AM
I've done a little more on the machine now.

Once I got the leadscrew set up and working on the x axis it became clear that running a single leadscrew down the middle of such a wide machine using skate bearings as linear rails was not a good idea. There was considerable yawing (pivoting) around the leadscrew, Ideally I should have run a drive system down each side, but I thought I'd try running a cable in a figure 8 path fixed to the outer sections of the cross beam which the leadscrew is attached to.

The cable had to be under tension to minimise any movement, I looked at varoius methods of tensioning the cable and decided to make a small winch which worked out very nicely. The main reason I went with a winch is that there is virtually no limit to the lenght of cable I could wind in to apply tension. Other methods were simpler but had very limited length capacity and took up too much space.

On the opposing mounting point I made an adjuster which I could use to square up the x and y axes. I used sliding door rollers in each corner for the cable to run in. The system appears to work ok, but not great, I may need to upgrade to a larger diameter cable but I'll wait and see how it performs in opperation first.

Here's some happy snaps...

Splint
01-04-2008, 08:53 AM
I folded up some trays for the power cables to run in across the top of the gantry and down the back edge of the side plate, that all worked out pretty well.

I also made some limit/home switch ramps and mounted the switches. I used an extension lead for mains power to the z axis as a die grinder will be used as the spindle and I'd like to be able to unplug it and take it out easily. I've also redesigned the motor mounts to include flex couplings. I initially didn't think I'd really need them but my design was causing the servo motor bearings to bind, so I went with flex couplings.

Greolt
01-04-2008, 04:24 PM
Looking good Splint.

When can we see it going??? :)

Greg

spidey4fun
02-10-2008, 04:51 AM
Very nice Splint... I got a small job waiting!!! hehehe :)

Brad.

Khalid
02-10-2008, 05:12 AM
goood job done splint...keep it up..and also share with us ur foundary practices:) would u??????

Splint
08-02-2008, 11:43 PM
Well things have progressed a little now. Vlad from Rutex is back on the scene again and I'm working through some issues with him which has caused some hold up for a while. When I've got the servo system sorted , I may need to change the energy chain over to a larger size as things are very cramped in the x axis chain. After that I'll need to finish installing the electrics in the control box (in the left of the pictures) and a few little cosmetic issues and I should be up and running.

Greolt
08-02-2008, 11:58 PM
Good to see the progress Splint.

Did I tell you about the Cable Chain supplier in Dandenong?

Greg

ynneb
08-03-2008, 12:25 AM
Dave, did Vlad give any indication that things are more stable. Is it safe to buy from him again?

Splint
08-03-2008, 03:17 AM
Thanks Greg, you did tell me but for the life of me I cant recall the name of the company, you might need to jog my memory.

Ynneb, I've been emailing Vlad and his future plans have not come up in conversation. I understand that the major componants from Rutex have been outsourced now so even if Vlad needs to head back to Europe the parts should still be available from Tom Eldridge. Unfortunately, Vlad being the designer of all the Rutex products, is the man with the most intimate knowledge of the products. Tom tries hard to help but has limited knowledge so if there's any really tricky problems arise you might be in trouble.

Greolt
08-03-2008, 03:42 AM
Here is the link and I sent you a PM

http://www.conductix.com.au/contacts.cfm

Greg

Splint
11-11-2008, 09:40 PM
Proof that it really does work..... I removed the spindle to give the oxy torch a go here. Seemed to work ok...


<object width="425" height="344"><param name="movie" value="http://www.youtube.com/v/TYNvVRlHR-E&hl=en&fs=1"></param><param name="allowFullScreen" value="true"></param><param name="allowscriptaccess" value="always"></param><embed src="http://www.youtube.com/v/TYNvVRlHR-E&hl=en&fs=1" type="application/x-shockwave-flash" allowscriptaccess="always" allowfullscreen="true" width="425" height="344"></embed></object>



<object width="425" height="344"><param name="movie" value="http://www.youtube.com/v/QhJjrgQ-pyU&hl=en&fs=1"></param><param name="allowFullScreen" value="true"></param><param name="allowscriptaccess" value="always"></param><embed src="http://www.youtube.com/v/QhJjrgQ-pyU&hl=en&fs=1" type="application/x-shockwave-flash" allowscriptaccess="always" allowfullscreen="true" width="425" height="344"></embed></object>

Greolt
11-11-2008, 09:53 PM
Good bit of adaptive thinking there Dave. :cheers:

Greg

spidey4fun
11-16-2008, 06:51 AM
Hey Dave.... so you've got those little bug sorted out?

Brad.

Splint
11-16-2008, 11:47 PM
Hi Brad,

yeah, it was the settings on the controller card, I had the upper amperage limit settings too low and I also had to turn off constant contouring. Seems to go pretty well now.

Cheers
Splint